e-brief August 31, 2010 FINANCIAL SERVICES Not Here? Housing Market Policy and the Risk of a Housing Bust By Jim MacGee Can a US-style housing bust happen in Canada? Recent swings in Canadian house prices have raised concerns about the possibility. To evaluate the likelihood of a USstyle housing market crash in Canada, the author examines what caused the US housing boom and bust from 2000 to 2009. A decline in underwriting standards played an essential role in the US housing market boom and dramatic bust. While monetary policy was very similar in both countries from 2000 to 2008, housing markets (especially the subprime component) were structured and regulated differently in each. Canadian housing policies, which avoided the sharp decline in underwriting standards seen in the US, worked well in reducing the possibility of a housing bust in Canada during 2008-2009, and continue to mitigate the risk of a massive wave of defaults in the future. Recent swings in Canadian house prices have raised concerns that a US-style housing bust looms on the horizon. 1 A comparison of housing market policies in Canada versus the US, however, suggests that there is little likelihood of a US-style surge in foreclosures or a collapse of house prices in Canada. Many of the concerns about the Canadian housing market are motivated by recent US experiences. Over the years 2000 to 2006, US prices appreciated nearly twice as much as Canadian houses (Figure 1 plots the United States S&P Case-Shiller 20 city composite index and the Canadian Teranet-National Bank 6 city composite index). 2 This rapid appreciation has been followed by an equally rapid decline, as US house prices declined by over 30 percent between 2006 and 2009, before staging a modest recovery late in 2009, only to fall into the doldrums again this year. The decline in Canadian house prices lagged the US, and was more muted, as house prices continued to appreciate until late 2008, before declining by roughly 9 percent between August 2008 and April 2009. This decline was followed by a rapid bounce back, with house prices returning to their pre-recession high by the end of 2009. After another price surge in the spring of this year, buyers and sellers remain on tenterhooks about the future path of prices. To evaluate the likelihood of a US-style housing market crash in Canada, one first needs to understand what caused the US housing boom and bust. As argued elsewhere (MacGee 2009), the Canada-US comparison The author would like to thank Finn Poschmann and Philippe Bergevin for helpful comments, as well as Ben Dachis, David Laidler, Thor Koeppl, Angelo Melino, Angela Redish, Nick Rowe and anonymous reviewers at the Bank of Canada and the CMHC. 1 See, for example, Wolf and Kwan 2008. 2 Both series are based on repeat sales, making these series a closer approximation to a constant-quality price index of nominal home prices than average house price sales.
suggests that a decline in underwriting standards played an essential role in the US housing market boom and dramatic bust. While monetary policy was very similar in both countries from 2000 to 2008, housing markets (especially the subprime component) were structured and regulated somewhat differently in each. Unlike in the US, the Canadian subprime market never expanded significantly into newer products, such as interest-only or negative-amortization mortgages, whose popularity grew rapidly in the US from 2003 to 2006. 3 Moreover, while subprime lending increased rapidly in both countries over 2000 to 2006, the Canadian subprime market remains much smaller than that in the US, as subprime lenders accounted for roughly 5 percent of mortgage originations in 2006 compared to 22 percent in the US (Mortgage Architects 2007). Mortgage delinquencies support the key role of underwriting standards in what transpired. While mortgage delinquencies began to climb before the recession in the US, they only began to rise in Canada after the economic slowdown began. Moreover, the decline in Canadian house prices between August 2008 and April 2009 did not result in a large increase in mortgage delinquencies. Government housing market policies likely played an important role in the evolution of differential underwriting standards. Governments in both countries provide guarantees to lenders against the risk of mortgage defaults for a significant share of mortgage lending. Government insurance of default risk creates what economists term moral hazard, since the insurance reduces the cost to lenders of making high-risk loans. This means that an essential part of government mortgage insurance programs is minimum underwriting standards, which reduce the incentive for banks to lower underwriting standards to generate loan volume. If these standards are too loose, then lenders have an incentive to make a large number of risky loans, which in turn sets the stage for high levels of mortgage defaults. Compared to the US, Canada has maintained tighter conditions on government backstopped insurance against mortgage default. The Bank Act requires all federally regulated institutions to purchase mortgage insurance on loans with down payments of less than 20 percent (25 percent before 2007). 4 This mortgage insurance is explicitly backstopped by the federal government, as the federal government guarantees the insurance provided both by the Canadian Mortgage and Housing Corporation (CMHC) and its private market competitors. 5 All high loan-to-value (LTV) mortgages (greater than 80 percent) insured through this program must satisfy underwriting standards that specify maximum loan-to-value ratios, maximum amortization periods, and minimum standards for credit worthiness and mortgage documentation. During the US housing boom, there was debate over relaxing these underwriting standards so as to allow Canadian banks to offer products such as interest-only loans and longer amortization periods. However, standards were only gradually relaxed to allow some forms of interest-only mortgages and longer amortization periods between 2003 and 2006. Subsequently, underwriting standards were tightened in July 2008, and again in April of this year. As a result, the government-backed insurance program has a small exposure to higher-risk loans. In contrast to Canada, prior to 2007, the US government directly guaranteed a small fraction of high LTV loans, primarily through the Federal Housing Authority (FHA). However, the US government indirectly provided insurance against default risk to a large fraction of risky high LTV loans through the Federal National Mortgage (Fannie Mae) and the Federal Home Mortgage Corporation (Freddie Mac), which purchased and guaranteed large groups of mortgages. These government-sponsored enterprises (GSEs) were viewed as having an implicit government guarantee to make good on 3 Interest-only mortgages grew from 2 percent of US mortgage originations in 2004 to 18 percent in 2006, while the market share of option ARMs (negative-amortization mortgages) rose from 5 to 9 percent of originations (Financial Crisis Inquiry Commission 2010). 4 Non-federally regulated institutions such as credit unions can also participate in the program, provided they meet the underwriting standards. Non-bank institutions can provide mortgage products (such as interest-only loans) that are not eligible for government-backed insurance. 5 The federal government is ultimately responsible for the obligations of the CMHC because of its status as a Crown corporation. The federal government also explicitly backs private mortgage insurers in the event that they could not meet their obligations to lenders, subject to a maximum of 90 percent of the original principal amount of a loan. 2 e-brief
Figure 1: House Prices in Canada and the US 220 200 (January 2000 = 100) 180 160 140 120 100 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 US Canada Source: The US series is the Case-Shiller 20 city composite. The Canadian series is the Teranet-National Bank 6 city composite. The value of each series is normalized to 100 at January 2000. mortgage-backed securities they guaranteed or owned. This implicit guarantee became explicit when the US government placed both institutions into conservatorship in September 2008. Both of these GSEs lowered underwriting standards before and during the US housing boom, becoming willing to purchase and guarantee mortgage-backed securities based on interest-only loans and negative amortization loans. One recent estimate is that the large portfolios of Fannie and Freddie and the FHA helped finance over 50 percent of subprime lending during the US housing boom (Pinto 2009). The private-sector suppliers of insurance against mortgage defaults in the US also accommodated lower underwriting standards. Although mortgage insurance is typically not required on private mortgages in the US, some form of insurance against default is generally sought by lenders, especially for securitized loans. By providing guarantees against mortgagebacked securities, private monoline insurers (such as Ambac Assurance and MBIA) effectively provided a significant amount of insurance against the risk of mortgage default, especially in the Alt-A loan (i.e., riskier than prime but less risky than subprime) and non-conforming loan segments. Ultimately, several of these monoline insurers essentially defaulted on e-brief /3
part of these insurance liabilities. These differences in government regulation and provision of mortgage default insurance are important, both in understanding the US housing market crash and the risks facing the Canadian housing market. 6 During the US housing boom, both private insurers and government-sponsored enterprises facilitated looser underwriting standards by increasing their exposure to high-risk mortgage products such as low-documentation, interest-only and adjustable rate mortgages. This build-up was a key factor in the US housing market crash, since when house price appreciation slowed (and then reversed) many borrowers found themselves unable to meet their monthly mortgage bills or to rely on higher prices to refinance. The resulting rapid rise in foreclosures, in turn, seems to have played a role in driving the large decline in US house prices since 2006. 7 This difference in government policy has helped to discourage the build-up in Canada of a large number of high-risk mortgage loans. In addition, the US crisis has made it more costly for many non-bank lenders to raise funds. As a result, many of these non-bank lenders have ceased to underwrite subprime loans in Canada. The small number of high-risk loans underwritten suggests that the modest decline in Canadian house prices predicted for 2011 (TD Economics 2010) is very unlikely to trigger a US-style surge in foreclosures. Conclusion Canadian housing policies, which avoided the sharp decline in underwriting standards seen in the US, worked well in reducing the possibility of a housing bust in Canada during 2008-2009, and continue to mitigate the risk of a massive wave of defaults in the future. To the extent that current policies impose on taxpayers a significant exposure to mortgage insurance guarantees and, therefore, some of the aggregate risk of a decline in housing prices, it will be in the interest of all Canadians if policymakers recall the lessons of the 2008-2009 experience should pressures to relax underwriting standards reoccur in the future. 6 A frequently cited difference that potentially also lowers the risk of a US-style housing market crash in Canada is the difference in the tax treatment of mortgage interest (whether mortgage interest is tax deductible). Unlike in Canada, mortgage interest is tax deductible in the US. A common argument is that by lowering the after-tax cost of mortgage debt, mortgage-interest deductibility leads to higher loan-tovalue ratios in the US since paying down mortgage debt is less attractive. Moreover, mortgage-interest deductibility may make some mortgage products, such as interest-only mortgages, relatively more attractive. However, there are two reasons to suspect that mortgageinterest deductibility may not have been a major factor in the US housing boom and bust. First, the different tax treatment of mortgage interest predates the recent housing boom and bust, and by itself should have a level, not a growth rate effect (i.e., it should have been capitalized in the level of house prices before the house-price boom). Secondly, the difference in mortgage-interest deductibility only matters for a subset of borrowers: mainly low net worth households with sufficiently high income to make itemization worthwhile. Many (mainly lower and lower middle income) US households find that it is not worth itemizing in which case they receive the same tax treatment as Canadian borrowers. Moreover, as Gervais and Pandey (2008) point out, households with sufficiently high net worth in Canada can effectively replicate the tax effects of mortgage interest deductibility by re-organizing their portfolios, since interest on investment loans secured by housing equity is tax deductible. 7 Another common argument is that legal differences make defaulting on mortgage debt more costly in Canada than in the US. Central to this view is the question of whether lenders have recourse to pursue a borrower who defaults for the difference between the mortgage balance and the sale price of a foreclosed home (referred to as deficiency judgments). In some US states (notably California and Arizona), deficiency judgments on principal mortgages are not permitted, while in most Canadian provinces lenders can pursue deficiency judgments. There is evidence that differences across US states in the availability of deficiency judgments, as well as the cost of foreclosure, matter for mortgage default decisions, especially for higher income (and net worth) households (Ghent and Kudlyak 2009). However, many US states (including Florida and Nevada) that permit deficiency judgments also experienced a housing boom and bust, along with high levels of foreclosure. Overall, the cross-state evidence suggests that while differences in recourse do matter for housing markets, they may not have played a key role in the US housing bust. 4 e-brief
References Financial Crisis Inquiry Commission. 2010. Preliminary Staff Report: The Mortgage Crisis. http://www.fcic.gov/reports/pdfs/2010-0407-preliminary_staff_report_-_securitization_and_the_mortgage_crisis.pdf, accessed July 29, 2010. Gervais, Martin, and Manish Pandey. 2008. Who Cares about Mortgage Interest Deductibility? Canadian Public Policy, 34(1): 1-23, March. Ghent, Andra, and Marianna Kudlyak. 2009. Recourse and Residential Mortgage Default: Theory and Evidence from US States. Federal Reserve Bank of Richmond Working Paper 09-10. MacGee, James. 2009. Why Didn t Canada s Housing Market Go Bust? Federal Reserve Bank of Cleveland Economic Commentary, September. Mortgage Architects. 2007. Comparing the Canadian and US subprime and alternative mortgage markets: Why the US subprime fallout is a USonly phenomenon. http://files.newswire.ca/40/masubprime.pdf, accessed: June 15, 2009. Pinto, Edward. 2009. High LTV, Subprime and Alt-A Originations Over the Period 1992-2007 and Fannie, Freddie, FHA and VA s Role. http://www.aei.org/doclib/pinto-high-ltv-subprime-alt-a.pdf, accessed July 28, 2010. TD Economics. 2010. Observation: Canada to Experience a Modest Home Price Correction in 2011. http://www.td.com/economics/special/pg0510_home_sales.pdf, accessed July 28, 2010. Wolf, David D., and Carolyn Kwan. 2008. The Tipping Point? Merrill Lynch Canada, Economic Commentary, September. This e-brief is a publication of the C.D. Howe Institute. Jim MacGee is Associate Professor of Economics at The University of Western Ontario. For more information call 416-865-1904. This e-brief is available at www.cdhowe.org. Permission is granted to reprint this text if the content is not altered and proper attribution is provided. e-brief /5